Google is an Artificial Intelligence Company

Google is search – right?

It has a big database of websites and other internet documents and is fantastic at letting you find what you are looking for within it.  It spends a lot of money crawling the web and on developing sophisticated algorithms that make your searches more and more relevant.  It also makes a lot of money from selling advertising on those searches.

If you check out the company website, Google’s mission is to “organize the world’s information and make it universally accessible and useful.”

And then I read “In the Plex” by Steven Levy.

This is a really great book and provides amazing insight into how Google was founded, how it works internally and even covers many of the problems it has had (China, Social Networking, etc).  But the most amazing realisation I had from reading it was that Google is not a search company.. Google is an Artificial Intelligence Company.  Search is just a means to an end.

The Google founders were always interested in how technology (algorithms) could augment human life – and then they got distracted by building a search engine.  Or did they?

Today, the Google index contains just about everything that is available online; and over the past 15 years they have created more and more sophisticated algorithms to better match questions (searches) with answers (web pages, documents, images).  The Google search engine is no longer simply a database with some nice search capability…  it is a brain.  And more accurately, it has become an extension of your brain – when was the last time you stopped to think before Googling the answer?

So this “brain” contains everything that has ever been digitised and over the past decade has been getting “trained” by us.  Every time we perform a search and then click on a result.. we are teaching it what the correct answer to the question was.  We have been doing this without fully realising it for a very long time.

The search engine has been the first step in Brin and Page’s dream of creating technology to augment human life.  But it doesn’t stop with the search engine.

The Google brain understands the spoken word too..  Google believed that speech recognition was a data problem – it needed massive amounts of examples (different “phonemes” and different tones, accents, local dialects, etc).  And so in 2007 it launched Goog-411 in the US.  It was a free directory enquiries line that recorded every request for information (about 3 billion per year) and was used by its engineers to create the speech to text capability for closed captioning in YouTube.. and eventually in other products (Voice Search, etc).

The Google brain understands location, distance, and topography.  Google maps with StreetView is amazing and also has many layers of AI engineering.  Faces can be recognised.  House numbers identified.  Every time you use Google Maps you are providing feedback to the neural network – speed of travel (are there traffic delays?), new routes, etc.  Google doesn’t just store some static digital maps – it has spatial awareness.

The Google brain knows who you are and who your friends are.  Although a late arrival into the social networking space — Google is all about algorithms and data and so the rise of social fell initially outside Google’s field of vision.  But now Google+ has arrived and we are all being forced to create an account if we want to access all that Google has to offer.  In this process, we are teaching Google more about us – our preferences, our habits, our interests both personal and professional, and who we know.  This is invaluable information for a company that derives its revenues from advertising, but it is pure gold for the AI engineers too.

But this post is not about “skynet” or scaremongering about some sinister plan for Google to take over our lives and the world.. I’m just sharing this eye-opening moment with you – that Google is not just a search engine.. the search engine is part of a larger vision.  That the search engine was the first step in an augmented life.  I’m not 100% sure if this was the plan from day one, but it certainly is now.  And – it suddenly means that Google projects that I thought were cool but irrelevant make more sense now.

The self-driving car for example.  I love this idea – I want one.. I want everyone on the road to have one.. but this would not have been possible without the AI that Google has developed over the past 15 years.  The maps and spatial awareness, the voice recognition and more importantly – the amazing machine learning that Google technology is capable of.  The car will not just be able to take you from point A to point B – it will remember your daily routes, learn where most traffic jams occur and check traffic monitoring services to find a detour if there are any congestions.  The same is true for Google Glass.

Google is an AI company; search was just the beginning and it seems clear now that AI was not an after thought.

If you want a further example of this strategy, let’s look at Ray Kurzweil.  He was appointed as Director of Engineering at Google in 2012 and will have full access to the Google brain to propel the company even further into Artificial Intelligence.  Here’s a quote from wikipedia about him:

Kurzweil is generally recognized as a public advocate for the futurist and transhumanist movements, due to his stances on life extension technologies, his efforts to forecast future advances in technology, and his interest in the concept of the technological singularity

What all this makes me realise – is that Apple have no chance in competing with Google with Siri.  Not unless they start investing heavily (e.g. buy/build a search engine, etc) in developing a competing platform for AI.

These are exciting times we live in.

 

One comment

Comments are closed.